Recently I've tried some of the desktop search products available for Windows, here are my considerations:
- probably better not to install 2 at the same time, they will try to index each other's temp files (I saw Copernic Desktop Search indexing some HTML files in a temporary folder created by one of Windows Desktop Search filters)
- I wasn't able to make Copernic Desktop Search and Windows Desktop Search to exclude folders matching a specific name pattern to be indexed. For example, I'd like to exclude any folder with "jspbuild" or "archive-cache" as I have no need to see duplicate search results or to manually filter them. And I'd even like to be able to tell them to scan for files with a specific extension but only if the containing folder or the filename matches a specific pattern (for example, scan .zip files but only if the name contains 'config')
- I didn't see a way to configure them to scan inside jar files, that is to treat them as zip files. Though, as Will seems to imply, I could replace any config.jar file with a config.jar directory containing the jar's files and ATG should not notice
- Windows Desktop Search is unbelievably slow at building its index. I deinstalled because after 2 days it still had not indexed the drives on data (my desktop machine with something like 120GB of used drive space, not small but not every single file did need to be indexed)
- Copernic Desktop Search I deinstalled too, it tends to grab 100% CPU time even when not reindexing and it drained the battery on libra (my laptop with a tiny 30GB drive). If Copernic people can fix this, it could be the best tool on the market
- For the moment I've settled on Wilbur: the interface is less visually appealing than those of its competitors, indexing PDF files requires some tinkering with pdftotext and gzip, but it's free, unobtrusive, fast on searches and not too resource-demanding when rebuilding its index
Copernic has been updated, and no longer tries to index anything if it‘s on a battery. Dunno about the 100% problem.